Advanced Learning Analytics Methods

This open-access book is the companion volume to *Learning Analytics Methods and Tutorials*, taking readers beyond the foundational methods into the most advanced and rapidly evolving areas of the field. Organized around four pillars — explainable AI, large language models, complex systems, and precision education — it introduces methods that are reshaping how we understand learning at both the collective and individual level. The first section tackles explainable AI and machine learning head-on: not just how to build predictive models, but how to make their decisions transparent and actionable for educators. Chapters cover classification, regularization, and both global and local explainability techniques, providing the tools to move AI from black-box predictions to interpretable insights. The second section addresses the emergence of large language models in education, with tutorials on natural language processing, BERT-based discourse coding, and using LLMs to generate automated, explainable feedback. The third section introduces complex systems thinking into learning analytics. Three dedicated chapters on Transition Network Analysis (TNA) — including frequency-based and cluster-based approaches — fill a critical gap in modeling how learning processes unfold over time. Complementary chapters on exploratory graph analysis and recurrence quantification analysis provide additional tools for capturing temporal complexity. The final section pioneers precision education: person-centered and idiographic methods that move beyond group averages to understand each learner as a unique dynamic system, including vector autoregression, unified structural equation models, and automated machine learning approaches. Each chapter follows the same proven format: accessible introductions to the underlying methodology, followed by step-by-step R tutorials with real datasets and reproducible code.

Table of Contents